Review Article
Guidelines for the investigation and management of a reduced level
of consciousness in children: implications for clinical biochemistry
laboratories
R Bowker1, A Green2 and J R Bonham3
Abstract
Addresses
1
Academic Department of Child Health,
Queen’s Medical Centre, Nottingham;
2
Department of Clinical Chemistry,
Birmingham Children’s Hospital,
Birmingham;
3
Department of Clinical Chemistry,
Sheffield Children’s (NHS) Foundation
Trust, Sheffield S10 2TH, UK
Correspondence
Dr J R Bonham
Email: jim.bonham@sch.nhs.uk
This article was prepared at the invitation of
the Clinical Sciences Reviews Committee of
the Association for Clinical Biochemistry
Whenever a child presents to hospital with a reduced level of consciousness,
admitting clinicians have to decide the underlying cause rapidly so that the correct
emergency treatment can be initiated. Unfortunately, the clinical presentations of
many of the possible diagnoses are very similar. The diagnosis often results from
investigations within the clinical biochemistry laboratory. In the past, clinicians have
had limited guidance on which tests to request when presented with a child with a
reduced level of consciousness. Guidelines have recently been developed relating
to all aspects of management of the child in a coma. Due to a lack of evidence in
the literature regarding the most appropriate first line tests for children with a
reduced level of consciousness, a formal consensus process (‘Delphi consensus’)
was performed using a large multidisciplinary panel of experts. The recommendations reached by this process include the list of initial (‘core’) tests to request for
all children with a reduced level of consciousness (excluding those immediately
after suffering a convulsion and those involved in obvious trauma). Depending upon
the results of these ‘core’ tests and the clinical condition of the child, further tests
may be requested later. The key point is that all the samples have been taken at
the time of presentation to provide the best chance of reaching a diagnosis and
correctly treating the child. The article reviews the recommended core investigations
and further tests and discusses how individual laboratories can help to implement
the guidelines jointly with their Emergency and Paediatric Departments.
Ann Clin Biochem 2007; 44: 506–511
Introduction
One of the greatest clinical challenges a paediatrician
or emergency physician faces is how to manage a child
who presents with an acute reduced level of consciousness, re£ected in a score of less than 15 on the modi¢ed
Glasgow coma score (GCSo15), or is only responsive to
voice, pain, or is unresponsive on the Alert,Voice, Pain,
Unresponsive (AVPU) scale, which scores the consciousness of patients as alert, responding to voice,
responding to pain or being unresponsive.
About 30 children in every 100,000 per year present
to hospital with a reduced level of consciousness not
due to trauma (‘non-traumatic coma’).1 In this group,
there is a 40% mortality rate.1 The list of possible
causes of non-traumatic coma is very long but the most
common are infection, drug intoxication, seizures and
inherited metabolic diseases.
506
Most of the conditions which cause a reduced level of
consciousness in children are clinically very di⁄cult to
di¡erentiate from one another, as the symptoms and
signs overlap. Clinicians therefore rely heavily on
laboratory information to reduce the di¡erential diagnosis list and focus appropriate treatment strategies.
Up until now there has been no expert advice on which
tests to request when a child presents with a reduced
conscious level. For example, the chance of diagnosing
the cause of a rare encephalopathy has been largely
dependent on the experience of the reviewing doctor
and the liaison and input from the laboratory.
Rare metabolic conditions by their very nature are
not encountered frequently by paediatricians or emergency department sta¡. These conditions may not even
enter the di¡erential diagnosis list, when investigations
are being considered. There is anecdotal evidence that
delays in diagnosing rare causes of a reduced level of
r 2007 The Association for Clinical Biochemistry
Reduced level of consciousness in children
consciousness are commonplace. Such delays in diagnosis can have devastating consequences for the child.
Providing advice on the initial investigations to perform in children with a reduced level of consciousness
should reduce the diagnostic delays and for the children with rare metabolic conditions improve their outcome.
An evidence-based guideline has now been developed for UK practice.2 Where evidence was not available a formal consensus process was used to help form
the recommendations. These recommendations
include:
the initial screening panel of tests which should be
performed on all children with a reduced conscious
level -- the ‘core investigations’;
the investigations to request if hyperammonaemia
is detected;
the investigations to request if hypoglycaemia is
detected;
the investigations to request if ketoacidosis without
hyperglycaemia is detected;
the investigations to request if no cause is detected
after reviewing the core investigations.
This review will summarize the main recommendations from the guidelines and speci¢cally focuses on
how the clinical biochemist and paediatrician can
work together to rapidly diagnose the treatable encephalopathies.
Core investigations
Once the decision has been made to investigate a child
for a reduced level of consciousness, the clinician faces
two challenges:
(1)
obtaining blood and urine and possibly cerebrospinal £uid or other samples from a sick child;
(2) prioritizing the tests to ensure the treatable
causes are detected promptly.
A large multi-professional UK panel involved in diagnosing and treating children with encephalopathies
have drawn up a list of core investigations2 which balance the need to capture all the major treatable causes,
without requiring an excessive amount of blood. The
core investigations are listed in Figure 1 as they appear
in the guideline. Due to di¡erences in laboratory practice across the National Health Service (NHS), clinicians will need local guidance from their clinical
chemistry department as to which bottles and what
volumes are required to perform these tests. The laboratory should provide speci¢c details.
All children
Perform the
following in
all children
with reduced
conscious
level except
those post
trauma and
those within
1 h post
convulsion
(see pages
4 and 5)
507
Capillary Glucose
Blood gas (capillary, venous, arterial)
Urinalysis (dipstick at bedside)
Laboratory glucose
(even if capillary glucose normal)
Urea and electrolytes (Na, K, Cr)
Liver function tests
Plasma ammonia
Full blood count
Blood culture
1−2 ml plasma
to be separated,
1−2 ml plain serum frozen and saved
10 ml urine to be frozen and saved
Figure 1 The core investigations for children with a reduced
level of consciousness
Hyperammonaemia
Analytical aspects
One test which requires special mention is plasma
ammonia. All hospitals with neonatal units, paediatric
wards and emergency admissions for children should
provide a robust and reliable service for measuring
blood/plasma ammonia 24 h a day, seven days a week.3
Laboratory sta¡ must be aware of the artifactual reasons for a falsely high ammonia concentration (see
Sampling errors below).
The majority of laboratories in the UK employ automated enzyme-based (glutamate dehydrogenase)
methods or dry slide chemistry for the measurement
of plasma ammonia. Re£ectance meters employing
dry chemistry strips for whole blood ammonia analysis
can be used for the initial assessment of patients with
suspected hyperammonaemia; however, the working
range of some of the older meters is limited and they
are therefore not suitable for the monitoring of hyperammonaemia when patients are being treated. Elevated results obtained by meters must be con¢rmed by
an alternative laboratory method.4 The laboratory has
a responsibility to make the clinician aware of this limitation. Use of such meters as point-of-care testing
devices on clinical units such as the emergency department are discouraged as they are prone to misuse
unless sta¡ are regularly trained and quality control
procedures are in place. Despite these concerns, the
results of a national audit recently conducted by
MetBioNet, a UK-based stakeholder alliance of 17
laboratories specializing in the investigation and diagnosis of inherited metabolic disorders, indicate that
these meters remain in widespread use (19 of 100
respondents) and this is a potential cause for concern
unless their operation is carefully regulated.
A raised ammonia concentration can be the ¢rst
suggestion that the patient has a metabolic encephalopathy. Several studies have shown that a better clinical
Ann Clin Biochem 2007; 44: 506–511
508
Bowker et al.
outcome is observed when the patient is exposed to
hyperammonaemia for as short a period as possible.5,6
The cut-o¡ for requiring active treatment for a raised
plasma ammonia is not clear cut,7 but most clinicians
would start treatment when the ammonia is greater
than 200 mmol/L.2
rated. It is important to warn clinicians about interpreting high ammonia concentrations if the sample
had not been free £owing when it was taken, before
initiating treatments and all elevated ammonia results
should be con¢rmed with a second specimen before
starting treatment.
Sampling errors
Causes of a raised ammonia
Erroneously high ammonia concentrations can occur
when the blood sample has been di⁄cult to take or has
been left to stand as whole blood at room temperature
for more than 15 min before the plasma has been sepa-
A raised ammonia may be caused by a number of di¡erent conditions, listed in Table 1. Further specialist tests
will be required to diagnose the underlying cause of the
hyperammonaemia. From the sample collected with
Table 1 Causes of pathologically raised plasma ammonia
Causes of hyperammonaemia
Relative NH3
concentration
Initial investigations
Inherited metabolic disorders
Defects of the urea cycle
N-acetylglutamate synthetase deficiency
Carbamyl phosphate synthetase deficiency
Ornithine transcarbamylase deficiency
Arginosuccinate synthetase deficiency
Arginosuccinate lyase deficiency
Arginase deficiency
þ
þ
þ
þ
þ
þþ
Urine orotic acid
Enzyme assay in liver or
Plasma and urinary
skin biopsy tissue
amino acids and urinary
depending upon the
disorder
organic acids
DNA analysis in some
cases
þ
þþ
þþþ
þþ
Urinary organic acids
and amino acids
Blood acyl carnitine
profile
Enzyme assay.
DNA analysis in some
cases
Plasma free carnitine and
acylcarnitine profile
Urinary organic and
amino acids
Skin fibroblast flux assays
Specific enzyme assay
DNA analysis in some cases
Organic acidurias
Propionic acidaemia
Methylmalonic acidaemia
Isovaleric acidaemia
3-Hydroxy-3-methylglutaryl coenzyme A (CoA)
lyase deficiency
Multiple carboxylase deficiency
Disorders of fatty acid oxidation
Multiple acyl CoA dehydrogenase deficiency
Very long chain acyl CoA dehydrogenase
deficiency
Long chain hydroxyacyl CoA dehydrogenase
deficiency
Medium chain acyl CoA dehydrogenase
deficiency
Carnitine acyl carnitine translocase deficiency
Carnitine transporter deficiency
Carnitine palmitoyl transferase 1 deficiency
Other inherited metabolic disorders
Hyperammonaemia, hyperornithinaemia,
homocitrullinaemia syndrome
Lysinuric protein intolerance
Hyperinsulinism hyperammonaemia syndrome
Mitochondrial respiratory chain defects
Pyruvate dehydrogenase deficiency
Ann Clin Biochem 2007; 44: 506–511
Confirmatory test
þþþ
þ
þþ
þþ
þþþ
þþ
þþ
þþ
þ
þ
þ
þ
þ
Urinary orotate
Urinary organic and
amino acids
þ
Intermediary metabolites
þ / þ þ þ including lactate, free
fatty acids and
3-hydroxybutyrate
Insulin measurement
Muscle enzyme and
DNA studies in
mitochondrial disorders
Enzyme assay
in some cases
Reduced level of consciousness in children
509
Table 1 Continued
Causes of hyperammonaemia
Acquired hyperammonaemia
Illness/sepsis
Urinary tract infection with a urease-producing
organism
Ribavarin therapy
5-Fluorouracil toxicity
Portosystemic shunting
Haemorrhage in the gastrointestinal tract
Renal tubular acidosis type 1
Relative NH3
concentration
Initial investigations
Confirmatory test
þþ
Clinical assessment
Microbiology results and
lack of other metabolic
diagnoses
þ
Blood glucose and
liver enzyme assay
Abnormal liver biopsy and
lack of other metabolic
diagnoses
þ
þþþ
þþþ
þþþ
Drug/therapy
clinical history
Absence of other metabolic
causes
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
þ
Reye’s syndrome
Drugs
Valproate sodium
Chemotherapy
Parenteral nutrition
Ethanol with starvation
þ
Blood ammonia approximately 600–2000 mmol/L
Blood ammonia approximately 200–600 mmol/L
þþþ
Blood ammonia o200 mmol/L
þþ
the core investigations already sent to the laboratory,
the following investigations should be undertaken and
may require referral to a Regional Centre:
plasma acylcarnitines
plasma amino acids
urine amino acids
urine organic acids
urine orotic acid
Treatment
Treatments are available to reduce the concentration of
ammonia. These include prompt intervention with
intravenous glucose and insulin to induce anabolism
and reduce further proteinolysis together with the
pharmacological use of sodium benzoate or sodium
phenylbutyrate to form benzoyl glycine and phenylacetyl glutamine which are excreted thereby removing
waste nitrogen. Arginine supplementation is also frequently given as an anaplerotic compound in the urea
cycle. In severe cases continuous veno-venous haemo¢ltration may be the treatment of choice. The sooner
these treatments are started the better, therefore it is
vital that the clinician is made aware of an abnormally
high ammonia concentration as soon as it is detected.
Appropriate initial therapeutic interventions should be
immediately available in all hospitals with neonatal
units or paediatric wards undertaking emergency
admission to help reduce brain injury. Subsequent transfer to a Regional centre may be indicated to arrange de¢nitive treatment including continuous veno-venous
haemo¢ltration.
Hypoglycaemia
Hypoglycaemia is often one of the ¢rst diagnoses to be
considered in a child with a reduced level of consciousness. This is often established at the bedside using a
capillary sample and one of the many types of glucometer. These devices are less accurate and more
imprecise than standard laboratory methods. In particular, at low concentration they may overestimate true
glucose results when compared with standard laboratory methods.8 The national guideline recommends
that all children with a decreased level of consciousness should have a laboratory glucose sent to con¢rm
the bedside capillary result.2
Therapeutic correction of hypoglycaemia may
reduce counter-regulatory hormone concentrations
(adrenaline, noradrenaline, cortisol and glucagon)
and consequently make laboratory investigations such
Ann Clin Biochem 2007; 44: 506–511
510
Bowker et al.
as intermediary metabolites (lactate, free fatty acids
and 3-hydroxybutyrate) more di⁄cult to interpret and
the disturbances observed in urinary organic acid
patterns and blood acylcarnitine pro¢les less prominent. Therefore it is important that, if possible, all the
investigations are undertaken before the glucose infusion is started. All laboratories serving hospitals with
neonatal units or paediatric wards should establish a
protocol to investigate unexplained hypoglycaemia in
childhood and a useful summary of the causes and
investigation is available.9 It may be valid to repeat the
tests at a later date just before feeds or even, in rare
instances, in a fasted state, although this can be very
hazardous in a child with a predisposition to hypoglycaemia.
Ketoacidosis
Sick children often have a metabolic acidosis due to
poor cardiac function and resultant tissue hypoperfusion. If there are ketones in the urine, then that
child is in a state of catabolism. A catabolic state is
normal after a period of starvation, but it may also signify the presence of a decompensating inherited metabolic illness. The di⁄culty is deciding when to
investigate for these rarer causes of ketoacidosis.
While mild elevations of plasma lactate are not uncommon and may be related to sampling, an unexplained
elevated lactate, greater than 3.5 mmol/L should
prompt further metabolic investigations to exclude
an organic acidaemia, aminoacidopathy, fat oxidation
defect or mitochondrial enzyme defect. The lactate
concentration can be measured on the £uoride oxalate
sample sent with the core investigations, but more
frequently can be ascertained from blood gas
analysis. Elevated lactate results identi¢ed in this
way should be con¢rmed using standard laboratory
methods if systematic di¡erences exist or reference
ranges di¡er.
The further tests required when ketoacidosis is
detected in the core investigations are:
plasma lactate
plasma amino acids
urine amino acids
urine organic acids
plasma acylcarnitines
The de¢nitive tests to detect the underlying cause
may sometimes require a skin ¢broblast biopsy for
enzyme assay or a sample of muscle for mitochondrial
studies. A muscle biopsy should be considered if the
child’s condition deteriorates rapidly, so that a diagnosis
may be made for genetic counselling of the parents at a
later date if the child does not recover.
Ann Clin Biochem 2007; 44: 506–511
The initial treatment for signi¢cant ketoacidosis is to
induce anabolism. This is achieved with a high concentration glucose infusion and an insulin infusion if
hyperglycaemia develops. Once the underlying cause
is identi¢ed this may help identify more speci¢c treatment.
Cause unknown
In many circumstances, the cause of a reduced level of
consciousness remains unknown after completion of
the core tests. If the clinical picture, glucose, ammonia
and pH have not provided any clues or if there is evidence of raised intracranial pressure or an intracranial
abscess is suspected, it is advised that the patient is
treated for infection and further tests are requested,
including imaging studies of the head. An important
cause of coma is drug ingestion,1,9 be that accidental
in the toddler age group or deliberate in the older child.
The further laboratory tests to request at this stage
include urinary organic and amino acids and blood
acylcarnitine pro¢le to exclude an occult metabolic
cause.
Urine toxicology should rapidly identify the presence
of substances in the child which may cause a reduced
level of consciousness. The panel of tests should include
drugs of abuse and any relevant medicines which the
child may have encountered in the home. This will need
an accurate history to be taken from the parent or
carer.
Implementation of the guidelines
For clinical guidelines to have an impact on patient outcomes they have to be successfully adopted and used by
health-care professionals. In the case of children with a
reduced level of consciousness, the success of the
guideline depends on the appropriate investigations
being requested at presentation and this requires e¡ective collaboration between clinicians and laboratory
sta¡.
Although the investigations requested should be
uniform for each patient, there will be variation
between di¡erent laboratories in how the tests are conducted. The local requirements of each laboratory need
to be available to the clinician taking the blood sample.
Far too often, precious blood which a junior doctor has
struggled to obtain is rendered un-useable by being
sent in the wrong bottle. As junior doctors move
between hospitals as part of their training, it is easy to
see how such mistakes happen if they are not reminded
of the individual laboratories requirements.
A project being established by a network of clinical
laboratories, MetBioNet, which hopes to address this
by encouraging local laboratories to help create an
Reduced level of consciousness in children
investigation package for clinicians. The pack should
include:
Investigation request forms, appropriately ¢lled in
with the core investigations for biochemistry, haematology and microbiology.
The correct blood bottles required for the core
investigations.
Information on how much blood is required in each
bottle to enable meaningful analysis.
Information on how quickly these samples need to
reach the laboratory.
In situations where only limited sample volumes are
obtained, the priority for analysis will depend on the
results of the initial investigations, for instance if a urea
cycle disorder is considered, plasma amino-acid analysis may be a priority, whereas if hypoglycaemia is prominent and a fat oxidation defect is suspected
acylcarnitine analysis would take priority.
Once a laboratory ‘kit’ is established and available,
the clinical team can access this when an unconscious
child presents, sending o¡ the correct tests in the correct bottles without delay or confusion.
The cost of implementing the recommendations for
the core investigations is di⁄cult to quantify. Most
large hospitals have the ability to perform the core
investigations on site without the need to invest in
new technology. As the number of children who present with a reduced level of consciousness is relatively
small, these recommendations are unlikely to have any
direct sta⁄ng implications and the marginal costs of
performing the core investigations is small. Consequently, cost should not be a barrier to implementation
and the bene¢ts of potentially avoiding serious longterm neurological injury are profound.
The guideline has been endorsed by the Royal College of Paediatrics and Child Health and by the British
511
Association for Emergency Medicine and was developed with input from the Royal College of Pathologists
along with many other organizations. It is hoped that
by standardizing care patients will have improved outcomes. A national audit will assess the implementation
and use of the guideline and best practice implementation examples can be shared. The guideline is freely
available at www.nottingham.ac.uk/paediatric-guideline
References
1 Wong C, Forsyth R, Kelly T, Eyre J. Incidence, aetiology, and
outcome of non-traumatic coma: a population based study. Arch
Dis Child 2001; 84: 193–9
2 Bowker R, Stephenson T, Atkinson M, et al. The Paediatric
Accident and Emergency Research Group. Evidence-based guideline for the management of decreased conscious level in children.
2006. Available at www.nottingham.ac.uk/paediatric-guideline
3 Losty H. Guidelines for the investigation of hyperammonaemia for
inherited metabolic disorders (appendix). 2004. UK National
Metabolic Biochemistry Network. www.metbio.net
4 Anderson D, Pennock CA, Henderson MJ, Holton JB. The
ammonia checker. Ann Clin Biochem 1984; 21: 69
5 Fitzgerald J, Angelides A. The prognostic significance of peak
ammonia levels in Reye Syndrome. Pediatrics 1982; 70: 997–1000
6 Uchino T, Endo F, Matsuda I. Neurodevelopmental outcome of
long-term therapy of urea cycle disorders in Japan. J Inher Metab
Dis 1998; 21(Suppl. 1): 151–9
7 Tokatli A, Coskun T, Ozalp I. Fifteen years’ experience with 212
hyperammonaemic cases at a metabolic unit. J Inher Metab Dis
1991; 14: 698–706
8 Balion C, Grey V, Ismaila A, Blatz S, Seidlitz W. Screening for
hypoglycaemia in the neonatal intensive care unit (NICU) with the
Abbott PCx glucose meter. BMC Pediatr 2006; 6: 28
9 Bonham JR. The investigation of hypoglycaemia during childhood.
Ann Clin Biochem 1993; 30: 238–47
Accepted for publication 14 August 2007
Ann Clin Biochem 2007; 44: 506–511